Brain basis of sharing and transmitting representations of social world

Abstract

Social communication is a crucial element of human behavior. Every day we resolve conflicts, empathize with our peers, exchange opinions and observe other's behaviors. While the brain basis of these processes has been studied in single individuals, it remains unresolved how such complex patterns of social interaction are parsed in the brains of interacting humans. This thesis addresses the brain basis of social communication in three domains: motor actions, language and emotions. These represent the main channels of human communication, and are tightly linked between each other. We used functional magnetic resonance imaging to collect the data, naturalistic stimulation as an experimental design principle and pseudo-hyperscanning to address the interaction in the experiments. We developed a novel hypeclassification approach, which combined pattern classification with functional realignment of data to investigate the shared neural codes between interacting individuals.In the first study we compared the neural coupling across multiple observers during active simulation versus passive watching of naturalistic boxing match videos by computing the time-varying intersubject phase synchrony of multiple observers' brain activity. We have shown that shared perspective synchronized brain networks involved in action execution and observation. In the second study we adopted a novel hyperclassification approach to investigate shared neural codes between action execution and observation in two individuals. We successfully showed that observed actions can be classified using the model trained on actor's data. The results revealed that action observation and execution share neural information in the brains of two interacting individuals. In the third study we used pseudo-hyperscanning to investigate the neural "coupling" between individuals telling emotional stories, and listeners of these stories. We measured the synchronization of their brain activity time series and revealed that as the experienced emotions became stronger and more similar between speaker and listener, their neural synchronization in attentional, limbic, somatosensory and midline structures increased. The fourth and final study investigated contextual effects on naturalistic speech comprehension. By manipulating context for a narrative, we addressed functional connectivity in the brain of listeners. Results of this study showed increase in functional connectivity in linguistic, attentional and error monitoring brain networks when individuals successfully understood speech in presence of relevant context. These results provide evidence for significant role of intersubject neural synchronization and shared neural codes in social interaction. Such synchronization may provide a window into mind state of another individual and enhance one's ability to understand and predict behavior of others.